There is a general consensus that smoking in young adults aged 18 - 25 presents a significant public health concern. Latest national surveys show that 30-day prevalence rates in young adults surpassed all other age groups, at an estimated 32% ~ 40%. A growing number of tobacco cessation interventions use the internet and mobile technologies to deliver tobacco education and treatment for youth and young adults. However, to date, these intervention strategies have generally ignored the social network dynamics and their impact on young adults' smoking behavior. To address these issues, we propose to use cellular phones as sensors to automatically capture social network influences of smoking. We will sample non-daily smokers aged 18 - 25 years from the City College of New York. We will oversample college freshmen to capture smoking trajectories during transition to college. We plan to provide participants with unlocked cellular phones. A utility program pre-installed on the cellular phones will automatically and unobtrusively capture social network dynamics by sensing the users' movement (via the Global Positioning System microchip), proximity to other smokers in their social network (via Bluetooth) and communication (phone logs). Ecological Momentary Assessments of smoking will also be made by the participants pressing a few easy keystrokes after each cigarette. Additional assessments will be made to capture the circumstances of smoking. Because the phone logs contain social contacts in real-time, they provide an ecologically valid representation of who affects the participants' smoking and in what settings. We will evaluate the validity of the sensor logs against conventional self-reported social network measures in describing how changes in social network dynamics affect smoking behavior. We will also test three main hypotheses: (1) young adult non-daily smokers smoke more often if they are closely connected to other smokers in their social network; (2) social network members who have high network importance have a stronger influence over other people's smoking; and (3) young adult non-daily smokers have a higher risk of transitioning to tobacco addiction if, over time, they shift their network position and choose to become connected more closely with other more frequent smokers. The ultimate public health goal is to use this information to guide the design of social-network-based prevention and/or intervention strategies for young adult smokers to capitalize on the vast potential for integrating social networking into cessation interventions delivered via mobile telephone technology. The enhanced ecological social network assessment methodologies can be applied broadly to numerous other areas of young adults' health behaviors including nutrition, physical activity, alcohol consumption, and risky sexual behavior.